Game Development Reference
In-Depth Information
Even some of the examples in the previous chapters on utility were based on
some assumed premises that could have been further explored and defined with
MAUT. For instance:
Pascal was far to simplistic when he relegated “live as if God exists� and “live as
if God does not exist� to equal status. I'm sure that Mr. Bentham would have
been more than happy to point out to the conflicted Mr. Pascal that those two
choices carry a significant amount of complexity in and of themselves.
When we built our tower over the barracks, we assumed that the barracks was
important to defend. Why is it important? What if we already had three bar-
racks buildings and this was a fourth that we were building for insurance? What
if we were currently beating back our opponent so badly that the likelihood of
having our barracks attacked was nil?
When we decided to build our settler and his escort soldier first, we worked from
the premise that sending out a settler earlier was better than doing so later.
Why? How much better? Is this as important if we are
not
concerned with ex-
panding our empire? Why would we be concerned about expanding our empire?
Are two smaller cities worth more than one slightly larger one? How much more?
All of these factors could be defined in terms of a utility attribute. We would
then be able to score the factors appropriately and include them in our decision.
I
NTHE
G
AME
The Engagement Decision Revisited
In Chapter 3, we showed a brief example in which we took multiple criteria into
account to construct a decision about whether or not an agent should engage the
enemy. In this case, the seven categories that Bentham defined would not be appro-
priate. Instead, we identified eight factors that we thought we should consider.
Those were:
Agent's health
Enemy's health
Agent's weapon
Enemy's weapon
Number of enemies
Proximity to a leader
Proximity to an important location
Agent's “anger� level
